Evergreen ILS Website

IRC log for #evergreen, 2017-03-08

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
05:00 pinesol_green News from qatests: Test Success <http://testing.evergreen-ils.org/~live>
07:11 rjackson_isl joined #evergreen
07:22 Callender joined #evergreen
07:53 collum joined #evergreen
08:07 csharp is it expected behavior that Firefox cannot connect to websockets on current master/2.11? (error: Firefox can’t establish a connection to the server at wss://master-xenial:7682/o​srf-websocket-translator.)
08:07 csharp Chrome connects as expected
08:25 collum_ joined #evergreen
08:44 mmorgan joined #evergreen
08:45 kmlussier joined #evergreen
08:45 kmlussier left #evergreen
08:48 Enjabain joined #evergreen
08:50 bos20k joined #evergreen
09:02 jeff csharp: it's unlikely that you have a valid cert that Firefox trusts without manual action for "master-xenial", so you're likely running into this: https://bugzilla.mozilla.o​rg/show_bug.cgi?id=594502
09:02 pinesol_green Mozilla bug 594502 in General "Method for accepting certificates for secure WebSockets connections (wss://)" [Enhancement,Unconfirmed: ] - Assigned to nobody
09:03 jeff csharp: You'll likely have better success with a real cert, or by signing a cert with a CA cert of your own that you then manually add to the Firefox trust store.
09:04 jeff csharp: That said, if you had success with Firefox and wss:// connections to hosts using not-otherwise-trusted certs before and it stopped recently, I'd be interested to know more.
09:15 Dyrcona joined #evergreen
09:16 dbs jeff: i filed my branch and bug 1671112
09:16 pinesol_green Launchpad bug 1671112 in Evergreen "Sample users and addresses do not contain any Unicode (non-ASCII) characters" [Undecided,New] https://launchpad.net/bugs/1671112
09:17 jeff dbs++
09:17 dbs Might be useful for testing search & display of Unicode in the user/address forms in webby?
09:18 Dyrcona And other things.
09:19 jeff looks like i got as far as a summary for the bug i started last night: ``Unqualified references in evergreen.unaccent_and_squash lead to index creation failures with pg_restore''
09:22 csharp jeff: yeah, I have a self-signed cert, but I thought it worked before with self-signed certs
09:23 csharp I'll dig
09:23 yboston joined #evergreen
09:24 Dyrcona Why is it that nearly every time I login to a server it says *** System restart required *** even though it was restarted since the last time I installed updates?
09:24 Dyrcona Systemd, maybe?
09:26 dbs jeff++
09:26 dbs Dyrcona: I think it's just awareness that so many libraries are loaded dynamically that stopping and restarting all of the affected services is pretty much impossible to do correctly
09:28 Dyrcona But, why does it say that when it was rebooted after the last apt upgrade, hmm? Linux never used to do this, and non-systemd systems don't seem to ask for reboots at apparently random intervals.
09:28 Dyrcona I'm gonna blame systemd no mater what, apparently. :)
09:28 dbs apparently :)
09:29 dbs I *think* it's because apt update gets cronned on many systems, so it's aware that updates need to be applied (and then restart again)?
09:30 Dyrcona Hmm.. I didn't actually set this one up, maybe the person who did said to automatically install updates....
09:30 Dyrcona That makes sense, and I don't get to blame systemd.... :)
09:30 Dyrcona So, applying Occam's Razor.... :)
09:35 dbs heh
09:35 Dyrcona So, doing my 30th to 35th installation in the last 3 months, and I'm doing the edi webrick steps.
09:36 Dyrcona I wonder if it would be worth it to provide updated installation scripts for jessie, wheezy, and xenial?
09:37 Dyrcona Or, making a generic script that figures out where the ruby gems go.
09:40 Dyrcona That is, pending the inclusion of Lp 1373690
09:40 pinesol_green Launchpad bug 1373690 in Evergreen "Direct EDI generation for ACQ orders -- AKA kill ruby webrick" [Wishlist,New] https://launchpad.net/bugs/1373690
09:40 Dyrcona I'm tempted to try that in production.
09:41 csharp Dyrcona: I thought of updating the script to be more distro-accommodating, but thought we'd already be moving to the "new" way before it became necessary
09:41 csharp Dyrcona: also, everything but B&T was well-tested at the Hack-A-Way
09:41 csharp so probably good to go
09:42 Dyrcona Well, looks like the new way won't make it before 3.0.
09:42 Dyrcona 2.12-RC is tomorrow, yeah?
09:42 csharp right
09:42 csharp Dyrcona: do your libraries use B&T?
09:43 Dyrcona The main issue is copying the mappers to the right place.
09:43 Dyrcona csharp: Yes, I'm pretty surse someone does.
09:43 Dyrcona heh....
09:43 Dyrcona sure...
09:43 * csharp saw someone with the nick "freudianlisp" in #code4lib yesterday
09:44 Dyrcona Ha!
09:44 jvwoolf joined #evergreen
09:44 Dyrcona Been messing with elisp a bit lately.
09:46 Dyrcona I seem to have two gems directories, but I wager one is a link to the other.
09:50 Dyrcona Nope. /var/lib/gems and /usr/lib/ruby/gems are not links to each other. Maybe the gems under 2.3.0 are?
09:51 Dyrcona Nope. Guess they're both in the gems path.
09:51 * Dyrcona has seen things like that before, like with Perl and Python.
09:52 csharp Dyrcona: the way berick and I were testing at the hackaway was to do "./edi_order_pusher.pl --test-mode --po-id 227 > brodart_new.edi" then compare to the original acq.edi_message in the DB, in case you want to go towards the non-Ruby EDI option
09:53 Dyrcona csharp: Right. I recall. I was gonna help but had an emergency to deal with.
09:53 Dyrcona I spoke briefly with tspindler about putting it in production and he wants to wait a bit.
09:54 csharp Bmagic supplied an example or two of a B&T order, and it looked okay, but could definitely do with more tests with real-life data
09:54 Dyrcona When I get the time...ha! ha! ha!...I'll give testing a whirl with our data.
09:54 csharp heh - let me know where you find the time, cuz I need some too :-)
09:54 Dyrcona And, bshum corrected me about 2.12-rc. It is next Wednesday.
09:54 csharp ah
09:55 Dyrcona Dunno why I thought it was tomorrow.
09:55 Dyrcona It would be up to kmlussier to decide if it's too late to kill ruby webrick in 2.12.
09:56 Dyrcona I suppose if you, csharp, berick, and I all signed off on it, then she'd probably say yes.
09:56 bshum At least, it's what the calendar tells me: https://wiki.evergreen-ils.or​g/doku.php?id=calendar:start
09:56 Dyrcona I meant you (csharp)... i guess the commas are tricky.
09:57 Dyrcona Anyway, I should figure out which gems directory openils-mapper is in so I can replace the files.
10:00 berick csharp: self-signed certs work w/ FF but you have to click through the SSL warning on the websocket port
10:00 berick https://server:7682
10:01 berick chrome on the other hand only requires you click through the ssl warning on the regular https port
10:01 berick FF requires both
10:03 csharp berick: oh - cool - I'll do that
10:03 mmorgan1 joined #evergreen
10:03 berick Dyrcona: if you want to test the edi stuff, I'd be happy to rebase
10:04 csharp berick++
10:04 berick well, i'll go ahead and rebase anyway.  would love to continue progress there regardless
10:15 dbwells joined #evergreen
10:16 csharp berick: count me in for more testing/prototyping
10:16 Dyrcona berick: That's OK. I'll rebase if I need to. I might try it with my custom 2.11 branch also.
10:17 csharp we're also up for piloting this here as soon as it's up to feature-parity
10:17 Dyrcona We'll probably backport it.
10:18 berick now have working/user/berick/lp1373690-acq-edi-perl-attrs -- should be easy to backport.  it's currently just 3 commits.
10:18 berick i would def. recommend testing a variety of vendor PO's before switching over
10:19 Dyrcona I could probably test all of ours with a bit of Perl. :)
10:20 Dyrcona Maybe pick a date range and let it go.
10:20 berick if you have any local JEDI changes, they will have to be replicated in the edi-attrs
10:20 berick which is currently accessible only via direct db access
10:21 Dyrcona I'll have to check. I'm not sure that we do.
10:23 bos20k joined #evergreen
10:25 bos20k joined #evergreen
10:33 Dyrcona berick: Is there a reason to not combine the branches on Lp 1373690?
10:33 pinesol_green Launchpad bug 1373690 in Evergreen "Direct EDI generation for ACQ orders -- AKA kill ruby webrick" [Wishlist,New] https://launchpad.net/bugs/1373690
10:35 kmlussier joined #evergreen
10:36 * csharp experiments with the new reporter UI
10:38 berick Dyrcona: they are combined.  just marked the attr bug as a dupe just a few mintues ago, in fact
10:38 csharp hey - that works!
10:39 * csharp will incorporate that into his reports session
10:39 Dyrcona berick: So the edi-attrs branch has the kill webrick code in it?
10:39 kmlussier I'm updating the MassLNC community demo server and am getting this error after I run grunt all - http://pastebin.com/MEPygksG
10:39 * csharp wonders what it will do with existing templates
10:40 csharp @dessert add ngCookies
10:40 pinesol_green csharp: The operation succeeded.  Dessert #52 added.
10:40 kmlussier berick / csharp: I already mentioned this to Dyrcona, but I think it's too late to get the kill ruby stuff into 2.12. But I would love to see it merged soon after the release so that we don't miss it again for 3.0.
10:40 berick Dyrcona: yes, user/berick/lp1373690-acq-edi-perl-attrs has everything
10:40 csharp kmlussier: I agree that it's too late for 2.12
10:41 berick yeah, too disruptive this late in the process
10:41 berick it needs some time to simmer in master
10:41 csharp but I'm getting crazier in my old(er) age and I'm willing to pilot stuff
10:41 csharp the new hold targeter rox
10:43 kmlussier csharp: I'm going to be very disappointed when I'm looking for a dessert and end up with ngCookies. ;)
10:43 csharp ha!
10:44 csharp ngCookieMonster
10:44 kmlussier :D
10:45 Dyrcona Cookie Waits? Search it on Youtube. Thank me later.
10:47 kmlussier Dyrcona++
10:48 csharp Dyrcona++
10:48 Dyrcona You're welcome! :)
10:48 * csharp has never been able to get into Tom Waits
10:49 csharp Bob Dylan, though is one of my heroes, so it's not just about the voice
10:49 kmlussier @love Tom Waits
10:49 pinesol_green kmlussier: The operation succeeded.  kmlussier loves Tom Waits.
10:50 kmlussier We used to play Tom Waits in the background when we walked our daughter to sleep for her naps. For some strange reason, it worked.
10:52 csharp eww - old old reporter for cloning existing templates
10:53 csharp that's gonna hurt
10:55 bos20k joined #evergreen
11:00 bshum kmlussier: Confirmed, I see the same error on my latest build from master too
11:00 csharp are there any docs or anything for transitioning from old-to-new reporter?
11:01 kmlussier bshum: Ugh. I was hoping it was user error.
11:02 jeff okay that's weird. launchpad or i seem to have doubled my bug title: ``Unqualified references in evergreen.unaccent_and_squash lead to Unqualified references in evergreen.unaccent_and_squash lead to index creation failures with pg_restore index creation failures Edit''
11:02 kmlussier csharp: It doesn't look like anyone from DIG has gotten to reports yet. https://wiki.evergreen-ils.org/dok​u.php?id=evergreen-docs:webclient
11:03 jeff double checking, pinesol -- what say you for bug 1671150
11:03 pinesol_green Launchpad bug 1671150 in Evergreen "Unqualified references in evergreen.unaccent_and_squash lead to Unqualified references in evergreen.unaccent_and_squash lead to index creation failures with pg_restore index creation failures" [Undecided,New] https://launchpad.net/bugs/1671150 - Assigned to Jeff Godin (jgodin)
11:04 jeff how about now? bug 1671150
11:04 pinesol_green Launchpad bug 1671150 in Evergreen "Unqualified references in evergreen.unaccent_and_squash lead to index creation failures with pg_restore" [Undecided,New] https://launchpad.net/bugs/1671150 - Assigned to Jeff Godin (jgodin)
11:04 jeff better!
11:18 csharp miker: if you're around... have you experimented with opening "old" (version 4 or before) templates in the angularjs reports UI?
11:19 csharp it'd be nice to just regexp_replace 4 with 5 and have it work
11:20 csharp but that's probably not how it works :-/
11:21 Christineb joined #evergreen
11:40 bshum kmlussier: So I found that if I change the line in bower.json for "angular-cookies": "~1.5.8",
11:40 bshum To put ~ in front of 1.5.8
11:41 bshum Then remove bower_components (rm -r bower_components)
11:41 bshum then bower install
11:41 bshum Then grunt all
11:41 bshum Success, happy
11:41 bshum Instead of ^1.5.8
11:41 * kmlussier looks
11:42 * Dyrcona would have to look up what the difference is between ^ and ~ in that context.
11:43 bshum Yeah me too, I just guessed and things seemed happier
11:43 bshum We use ~ for most of the other stuff
11:43 bshum I think it means minimum
11:43 bshum Ah
11:43 bshum http://stackoverflow.com/questions/​22343224/whats-the-difference-betwe​en-tilde-and-caret-in-package-json
11:43 bshum The ^ matches on any first number
11:44 kmlussier bshum: Yup! That works for me. I can create a branch.
11:44 bshum So that's why we see the version 1.6.3 floating around in the error
11:44 kmlussier bshum++
11:44 bshum It's pulling in stuff that's too new by comparison
11:45 bshum By switching to ~, we grab the latest 1.5.x which is 1.5.11 same as the rest of the stuff
11:45 jeff ~1.2.3 locks you at 1.2.x, while ^1.2.3 locks you at 1.x.x
11:45 pinesol_green jeff: I am only a bot, please don't think I'm intelligent :)
11:47 Dyrcona jeff++ bshum++ kmlussier++
11:47 Dyrcona pinesol_green: Intelligence is relative.
11:47 pinesol_green Dyrcona: Try restarting apache.
11:47 pinesol_green I'll remember that, Dyrcona
11:47 Dyrcona ~intelligence
11:47 pinesol_green intelligence is relative.
11:47 Dyrcona :)
11:48 Dyrcona BTW, restarting apache didn't help.
11:48 jeff arguably the official docs try to be a bit more semantic about it:
11:48 jeff ~version "Approximately equivalent to version" See semver
11:48 pinesol_green jeff: I am only a bot, please don't think I'm intelligent :)
11:48 jeff ^version "Compatible with version" See semver
11:48 Jillianne joined #evergreen
11:48 Dyrcona pinesol_green: Version is Approximately equivalent to version.
11:48 pinesol_green Dyrcona: (version takes no arguments) -- Returns the version of the current bot.
11:48 pinesol_green I'll remember that, Dyrcona
11:48 jeff see https://docs.npmjs.com/misc/​semver#tilde-ranges-123-12-1 and https://docs.npmjs.com/misc/se​mver#caret-ranges-123-025-004
11:49 Dyrcona ~Version
11:49 pinesol_green version is Approximately equivalent to version.
11:49 Dyrcona :)
11:49 Dyrcona OK. I'm gonna get some lunch.
11:49 Dyrcona I'm at a good stopping point in what I'm doing.
11:49 Dyrcona ~lunch
11:49 pinesol_green Factoid 'lunch' not found
11:50 Dyrcona pinesol_green: Lunch is Lunch is good.
11:50 pinesol_green Dyrcona: Fire BAD! Reading GOOD!
11:50 pinesol_green I'll remember that, Dyrcona
12:00 mmorgan joined #evergreen
12:00 kmlussier In what release did we upgrade to Angular 1.5?
12:01 * kmlussier answers her own question. 2.11
12:04 kmlussier angular-cookies is in 2.11 too, so it looks like we probably should backport this one.
12:05 bshum Good point..
12:05 bshum Bug fixing, whee
12:06 khuckins_ joined #evergreen
12:08 * kmlussier disappears for the afternoon to focus on other things.
12:08 bshum Oooh progress bars are pretty :)
12:18 brahmina joined #evergreen
12:39 Bmagic @csharp: We still have that edi work outstanding. I have it on my list of things to follow up on. If I can lend some of my example data to the greater good, ping me
12:39 pinesol_green Bmagic: MARC still isn't dead yet, alas
12:39 Bmagic oops, IRC != slack
12:46 Dyrcona I always assumed it was because of twitter. :)
12:49 dbwells_ joined #evergreen
12:51 jeff Bmagic: thank goodness for that.
12:51 Bmagic :)
12:53 sandbergja joined #evergreen
12:53 Dyrcona :)
13:01 jeff Hrm. Do we have documented recommendations for pg_dump/pg_restore of an Evergreen database? I think the only place where pg_dump appears in docs is the draft 2.1 docs pdf...
13:09 dbwells joined #evergreen
13:27 Bmagic jeff++ # I agree, Postgres management documentation specific to Evergreen would be nice. I have some things to contribute to that
13:28 jvwoolf joined #evergreen
13:32 jeff I was starting to draft a test plan for a bugfix on bug 1671150 when I thought to look.
13:32 pinesol_green Launchpad bug 1671150 in Evergreen "Unqualified references in evergreen.unaccent_and_squash lead to index creation failures with pg_restore" [Undecided,In progress] https://launchpad.net/bugs/1671150 - Assigned to Jeff Godin (jgodin)
13:34 jeff I think common practice is "use create_database_extensions.sql to create the db and then restore into it", and even less documented is "set the search_path on the new database", and of course "make sure any users exist if this is a new/different cluster than where the dump originated..."
13:53 csharp @dunno add PHRASING!!!
13:53 pinesol_green csharp: The operation succeeded.  Dunno #53 added.
13:55 Dyrcona Ah, bummer. I just told dpearl he should look at that bug....
13:55 Dyrcona :)
13:58 jeff bummer?
14:04 Dyrcona I thought you were deciding it's a documentation issue....
14:05 Dyrcona Maybe I misunderstood.
14:05 Dyrcona Anyway, I've usually just pg_restore'd and it work'd. :)
14:05 Dyrcona I don't think I have that branch though.
14:09 jeff and if you're not stopping on errors with pg_restore, you get a working database, you just lack certain indexes.
14:10 Dyrcona I don't get any errors with pg_restore.
14:10 Dyrcona Not lately, just warning about columns not have a datatype. Think it's related to a custom view or table made from a query.
14:10 jeff is this an Evergreen 2.11 or higher database?
14:11 Dyrcona Nope. 2.10. :)
14:12 jeff that would pre-date these indexes. :-)
14:13 Dyrcona Well, there you have it.
14:14 Dyrcona jeff: Is it just these unaccent and squashed indexes or are you referring to other indexes?
14:14 * Dyrcona just wants to make sure.
14:17 jeff just the four unaccent ones on actor.usr
14:17 jeff family/first/second names, usrname
14:19 Dyrcona OK.
14:20 * Dyrcona dots his ts and crosses his is to make sure the server is really ready.... ;)
14:24 Dyrcona Good thing I did. I almost forgot about the smarthost config for email.
14:52 jeffdavis Is anyone using nginx as a proxy with subdomains for different OPAC skins?
14:52 jeffdavis I'm trying to set this up but with the default nginx osrf-ws-http-proxy config I always get the default skin, not sure how to adjust accordingly.
14:55 _bott_ left #evergreen
15:08 _bott_ joined #evergreen
15:27 rlefaive joined #evergreen
16:43 kmlussier joined #evergreen
17:01 pinesol_green News from qatests: Test Failure <http://testing.evergreen-ils.org/~live>
17:05 kmlussier Oh, good! We already have a fix for that!
17:09 mmorgan left #evergreen
17:18 csharp @blame pinesol_green
17:18 pinesol_green csharp: Your failure is now complete, itself.
17:23 Bmagic csharp: I haven't found you at code4lib yet
17:27 csharp Bmagic: I'm not there!  We decided to skip it this year. :-/
17:29 Bmagic ah! that would explain the lack of csharp in the hallways. Anyone in this channel here? I've ran into gmcharlt a couple of times
17:29 csharp Terran has been participating remotely - I tuned in yesterday, but not today
17:29 berick i was going to go, but opted out when they moved it.
17:29 * kmlussier has still never attended a code4lib.
17:29 csharp honestly, it's not really doing it for me this year - more about process/career stuff than code as far as I saw yesterday
17:30 csharp yeah, I would have gone to Chattanooga for sure
17:31 berick csharp: I had already started (mentally) planning my side-trip to Nashville :(
17:31 Bmagic csharp: I agree, many of the presentations are abstract
17:32 kmlussier Do they have a location for next year's conference yet?
17:33 Bmagic There was a lightening talk yesterday that had my ears perked up. SimplyE
17:33 csharp kmlussier: nothing posted here yet https://code4lib.org/conference
17:33 csharp Bmagic: one of our administrative librarians is very interested in that
17:33 berick 2nd call for proposals went out last week
17:33 Bmagic lightning*
17:34 berick kmlussier: i highly recommend going at least once.  I always like it.
17:34 csharp it would make sense if your library subscribes to multiple e-resource vendors (most of ours subscribe to one, maybe two)
17:34 pinesol_green [evergreen|Kathy Lussier] LP#1671165: Fix web client build error - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=7e46ea8>
17:34 csharp kmlussier: +1 to berick
17:34 csharp I've been, let's see... 5 times
17:34 kmlussier I should have gone the year it was in Providence. But I was involved in other things back then.
17:35 kmlussier I wish they would host it in Providence again. A five minute drive from my house. 10 minutes during rush hour.
17:35 berick yeah, raleigh was pretty great for me
17:35 Bmagic haha, L.A. - not possible to go anywhere in 5 minutes. The traffic out here is out of this world
17:35 berick although, i prefer to travel so I don't have to commute each day
17:36 csharp my favorite venues have been in NC - Asheville and Raleigh
17:36 csharp but all of them have been great
17:37 * csharp got to jam with other music-y folks last year
17:37 berick every year in my EG conference survey I say we should make the EG conf more like code4lib.  single track, 20-minute talks.  it's my favorite.  everyone in the same room.  no filler content.  broad range of topics.
17:38 berick i get that's probably not going to happen..
17:38 csharp berick: I like the format too - it might be worth trying out for EG too
17:38 Bmagic berick: I can see pros and cons to the single track
17:38 berick i think people would learn a lot
17:39 kmlussier Yeah, but code4lib is more likely to have an all-tech audience, isn't it? EG has a very broad range of technical abilities in the audience.
17:39 Bmagic kmlussier: I was trying to think of a way to say exactly that
17:39 berick by "filler content" i mean the need to fill 50 minutes of time w/ 20 minutes of actual information.  (which I sometimes find myself doing)
17:39 * kmlussier also isn't sure she's capable of limiting herself to 20 minutes when she talks about anything.
17:40 Bmagic Some presenters are doing 15 minutes, and some are doing 30
17:40 Bmagic I think you get to decide when you submit your proposal
17:40 berick kmlussier: maybe, re: tech audience, but a lot less true than it used to be.
17:40 csharp c4l seems to be trying to balance its original mission as a hard tech conference and inclusivity of newly (or non-)technical folks
17:41 csharp I've not been to a LITA conference, but what I'm seeing topic-wise looks more like that to me
17:41 csharp (this year)
17:42 berick Bmagic: oh, i kind of like that too
17:42 kmlussier LITA is one I've wanted to go to for years.
17:42 kmlussier How large is code4lib?
17:43 csharp kmlussier: 300+ usually
17:43 csharp used to be waaay smaller
17:43 kmlussier Wow!
17:43 csharp there was a cap in place until 2013
17:43 csharp they used to cap it at 250 to keep it intimate
17:51 * kmlussier puts her Mom/Taxi Driver hat on and departs for the evening.
17:51 kmlussier Have a nice night everyone!
17:55 Enjabain joined #evergreen
17:59 * csharp briefly has image of kmlussier as Travis Bickle in his head, then shakes it off
18:04 berick @band add Origin Hamster
18:04 pinesol_green berick: No, you're a puzzleheaded kraken!
19:54 Bmagic @later tell kmlussier attendance this year is 425
19:54 pinesol_green Bmagic: The operation succeeded.
19:55 rlefaive joined #evergreen
19:55 Jillianne joined #evergreen
20:18 jeff Bmagic: on the subject of "Postgres management documentation specific to Evergreen", what kinds of things did you have in mind?

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat